Ruixin Micro Release Android 8.1 Nnapi Sdk, Ai Operation Performance Greatly Improved

Set micronet messages, Rockchip formally releases the Android 8.1 Neural Networks API (NNAPI) optimized SDK based on the RK3399 platform, providing a more versatile and more powerful AI computing support for the model. It is suitable for various applications based on the mainstream model architecture, such as face recognition, ADAS, commodity identification, fatigue detection and so on.

RK3399 has the characteristics of high performance, high expansion and versatile application. CPU uses big.LITTLE size core architecture, dual core Cortex-A72+ four core Cortex-A53, and has a technical lead in overall performance and power consumption. GPU uses a new generation of four core ARM high end image processor Mali-T860, which integrates more bandwidth compression technology and has excellent overall performance. Android 8.1 can fully invoke neural network API (NNAPI) for hardware acceleration, so that the performance of RK3399's AI operation is greatly improved.

The release of RK3399 SDK is a GPU acceleration scheme based on Android 8.1 NNAPI, which provides accelerated support for AI related applications developed with Android NNAPI, with four advantages:

1. wide compatibility: Standard API, directly supporting all kinds of APK applications developed based on Android NNAPI.

2. versatility: it can support a large number of mainstream model architectures and can be applied to various applications based on mainstream model architecture, including face recognition, ADAS, commodity identification, and fatigue detection.

3. surge in performance: real-time performance can be achieved in multiple tasks, such as using MobileNet for image recognition, with the highest frame rate of 23.2 frames.

4. lower power consumption: Based on GPU efficient computation, full load power consumption is only 1W.
According to the APK test data of image recognition and target detection, the mainstream model performs well:

AI computing is in the period of explosion. Rockchip artificial intelligence chips have been widely used in the fields of image recognition, intelligent security, intelligent driving, voice recognition, consumer electronics and so on. This disclosure is based on RK3399 platform Android 8.1 NNAPI optimization SDK, will make its AI chip product line more hard power.